What Meridian did
Alex wrote the daily checklist on a shared doc in week one and required the biller to time-stamp each item. It felt like overkill at eight visits a day. In month four, at forty visits a day with two clinicians, the same checklist was the reason Meridian’s clean claim rate never dropped below 94%.Week 1: install the daily loop
Assign every item an owner by name, not by role.Morning (30 minutes)
- Pull tomorrow’s schedule; run eligibility (270/271) on every patient
- Flag coverage problems and have the front desk call those patients today
- Flag services needing prior authorization; start the requests
- Check the clearinghouse for overnight 999 / 277CA rejections
- Work every rejection to resubmission before touching anything else
Afternoon (60 minutes)
- Enter charges for yesterday’s completed visits
- Confirm every completed visit has either a charge or a documented reason it doesn’t
- Run the scrubber; clear every edit
- Submit the batch
- Post any 835s received; work the exception queue
- Confirm each deposit ties to its remittance via TRN
End of day (10 minutes)
- Record the day’s numbers: visits, charges entered, claims submitted, rejections, payments posted
- Note anything unresolved for tomorrow
Week 2: install the weekly loop
Pick a fixed day. Meridian used Tuesday.- Denial queue, triage everything new by CARC group; nothing older than 14 days untouched
- AR aging, review the 60+ bucket line by line, the 90+ bucket with the ops lead
- Credit balances, review the report; anything over 30 days gets resolved
- Credentialing grid, update statuses; call any payer silent for two weeks
- Patient balances, confirm statements went out on schedule
- Root-cause tagging, for each denial worked, record why it happened, not just how it was fixed
Week 3: build the dashboard
You need one page, updated weekly. Six numbers and one table.
Plus a table of denials by CARC, sorted by dollar value. That table tells you what to fix next, every week, without any judgment call.
Track each metric by payer as soon as you have more than one.
Week 4: close the prevention loop
The mechanism that separates practices that improve from practices that just keep working:1
Tag every denial with a root cause
Use a fixed taxonomy: eligibility, authorization, coding, credentialing, timely filing, coordination of benefits, documentation, demographic error. Free text defeats the purpose.
2
Rank causes by dollars, weekly
Not by count. Ten 900 denial.
3
Change one upstream process
One per week. If eligibility failures top the list, add a second verification at check-in. If authorization failures top it, build the auth-required list per payer into the scheduling workflow.
4
Verify next week
Did that cause drop? If not, the fix didn’t work — change something else rather than adding a second layer to a failed one.
Documenting who does what
Write it down, because your first biller will not be your only biller.
Keep the last row’s boundary clear. Billing staff are MSO employees; coding decisions belong to the PC. The MSO provides the people and the systems; it does not determine what diagnosis or procedure codes are assigned. California’s SB 351 names billing and coding among the functions a management entity may not control.1 See Run a CPOM self-audit.
What good looks like at day 30
- Zero unbilled encounters older than two business days
- Every 277CA rejection worked the day it arrives
- A denial queue with a named owner and nothing older than 14 days
- A weekly dashboard someone actually reads
- At least one upstream process changed because of denial data
Next
Your first denial
A CO-197 arrives. Walk it through end to end.
Sources
- Cal. S.B. 351 (2025), effective January 1, 2026. Summary: Quarles, California Cracks Down: New Laws Governing the Corporate Practice of Medicine and PE Deals.